Literature on Saccharopolyspora erythraea physiology can be discussed in two main topics and both are presented below. The first topic is the effect of media components on growth and erythromycin production, and the second is the effect o f fermentation parameters on growth and erythromycin production.
Ph .D . Thesis M e ta b o lic E n g in eerin g o f S. erythraeâ
3.2.1.5.1 Effect of Media Components on Growth and Erythromycin Production
The differences in the physiology of S. erythraea when grown on complex or defined carbon and nitrogen sources was observed by Smith et al. in 1962. They found that when S. erythraea was grown on a complex nitrogen source (soybean oil)
erythromycin biosynthesis exhibited a non-growth related secondary metabolite profile and production continued until the substrate was depleted. Addition of sucrose caused renewal o f erythromycin production, however, if nitrogen was added with sucrose then production was not renewed. It was presumed that the processes of growth consumed the intermediates, energy, or cofactors that would otherwise be used for erythromycin production. Using a defined nitrogen source (glycine) erythromycin production followed a growth related profile. They observed limited growth which may have been due to the inability to fully use media components and, therefore, allowed for simultaneous erythromycin synthesis. Addition o f sucrose and glycine caused renewal of both growth and erythromycin biosynthesis (Smith et a l, 1962). It was also suggested that pyruvate, generated internally from sucrose or glycine, was only available for erythromycin biosynthesis when it was not used for growth. The situations where pyruvate may not be needed for growth were nitrogen limitation or inhibitory concentration of erythromycin.
When S. erythraea was grown on glucose, a strong but temporary suppression of antibiotic formation was observed. During this antibiotic suppression phase, growth still occurred and the maximum erythromycin suppression occurred at 20 g/L glucose. Glucose only suppressed antibiotic formation when added before the
production phase (Escalante et a l, 1982). Glucose repression o f antibiotic production has also been observed in other Actinomycetes (Martin and Demain, 1978).
The effect of ammonium on erythromycin production was studied and it was found that total erythromycin inhibition was obtained with 100 mM ammonium chloride. Ammonium nitrate and ammonium sulfate also repressed erythromycin formation, however, organic nitrogen sources (glycine, leucine, glutamine) did not reduce erythromycin production. Addition of ammonium after the onset o f erythromycin synthesis inhibited production after a 6 hour delay (Flores and Sanchez, 1985). In a separate study, erythromycin production was higher when grown with ammonium nitrate than ammonium sulfate (Potvin and Peringer, 1994a). Erythromycin
Ph .D . Thesis M e ta b o lic E n g in eerin g o f S. e r y th n e z
production was growth related with ammonium sulfate and growth dissociated with ammonium nitrate. A hypothesis was proposed suggesting that glutamate produced from ammonium could be converted to glutamine via nitrate induced glutamine synthetase where glutamine could then be used as an erythromycin precursor (Potvin and Peringer, 1994b).
Erythromycin A production increased with growth rate in phosphate-limited defined medium chemostat culture. Erythronolide B, 3-a-L-mycarosyl-erythronolide B, erythromycin D, and erythromycin B+C followed this same relationship. The relative level of each compound decreased thorough the progression o f the synthesis reactions (see section 2.2.1.1), where erythronolide B had the greatest amount (~7
pmoles/gDCW/h), then 3-a-L-mycarosyl-erythronolide B (-5.5 pmoles/gDCW/h), erythromycin D and B+C (~3 pmoles/gDCW/h), and finally erythromycin A (~2 ^moles/gDCW/h) (Trilli et a l, 1987).
In different nutrient-limiting media it was observed that erythromycin production was growth related with nitrogen-limited medium and followed a classical secondary metabolite or growth dissociated pattern with carbon and phosphate-limited media (McDermott et a l, 1993). In a separate study, the onset of antibiotic synthesis was coordinated with the minimal protein synthesis rate and minimum ratio o f charged to uncharged tRNA. Antibiotic synthesis and increased uncharged tRNA are generally considered stringent responses and it was proposed that there may be some connection between the rate of protein translation and antibiotic synthesis (Wilson and Bushell,
1995). Bushell et a l (1997) extended these observations o f nutrient limitation and proposed a model suggesting that glucose is a high affinity limitation which means that uptake rates are high until substrate is exhausted. At this point the supply of ATP needed for amino acid activation becomes limiting causing the protein synthesis rate to decrease. For nitrate, which is considered a low affinity limitation, uptake rates are decreased early in growth cycle. The supply of amino acids becomes limiting and the protein synthesis rate decreases earlier in the growth cycle compared to glucose limited cultures. Once the protein synthesis rate decreases, the uncharged tRNA accumulates and binds to the ribosome. Guanosine tetraphosphate (ppGpp), which is
Ph .D . Thesis M e ta b o lic E n g in eerin g o f S. tryihm tZi
produced upon the reduction of the charged/uncharged tRNA ratio, may then initiate a cascade that results in antibiotic production (Bushell et a l, 1997a).
3.2.1.5.2 Effect of Fermentation Parameters on Erythromycin Production and Morphology
Oxygen supply is an important parameter in fermentation. It was found that erythromycin was produced in oxygen limited and oxygen sufficient cultures. The maximum biomass and erythromycin production were the same, although, growth rates and erythromycin production rates were reduced in the oxygen limited cultures (Clark et a l, 1995). In a separate study, the maximum biomass and erythromycin production were somewhat decreased when DOT was reduced from 65% to 10%. When the agitation rate was increased from 500 RPM to 750 RPM, the biomass and erythromycin production was further reduced presumably due to mycelial breakage (Heydarian et a l, 1996). In another study, it was shown that the amount o f initial rapeseed oil did not affect the growth rate or specific erythromycin production in a 2 L fermentation volume. However, at a 7 L fermentation volume, the growth was not affected but the specific erythromycin production increased with increased oil
concentration. It is believed that the oxygen limitation at the 2 L scale may have been masking the effect of the varying oil concentrations and that the 7 L fermentation shows the actual effect of increased oil concentration (Mirjalili et a l, 1999).
Erythromycin production was greater using an oil based medium compared to a soluble complex medium but required longer fermentation times and was more difficult to recover with membrane filtration technology. Overall, the increased erythromycin from the oil based medium in the feed stream to filtration did not benefit erythromycin recovery (Davies et a l, 2000). Erythromycin production was increased using a cyclic fed batch culture. Cyclic fed batch culture produced higher levels of erythromycin than chemostat culture. This was likely due to the high dilution rates typically used in chemostat cultures which do not simulate the low growth rates typical o f erythromycin production (Lynch and Bushell, 1995).
There has been several studies on the hyphal strength of S. erythraea and the hyphal size at which erythromycin is produced. It was found that hyphal fragments greater than 88 micron exhibit erythromycin production (Martin and Bushell, 1996). This finding was consistent for stirred flasks and for baffled or unbaffled shaken flasks
Ph.D . T h esis M e ta b o lic E n g in eerin g o f S. erythraea
(Bushell et a l, 1997b). S. erythraea mutants which showed decreased branching frequency also had increased hyphal strength. This increased hyphal strength increased the number o f particles with diameters greater than 88 micron and hence increased erythromycin production (Warded et a l, 2002). The strength of hyphae were found to be about 60% greater during growth phase compared to stationary phase. This is thought to be due to a thinner cell wall during stationary phase as the tensile strength was the same for both phases (Stocks and Thomas, 2001).
3.2.1.5.3 Red Pigments Produced by Saccharopolyspora erythraea
Saccharopolyspora erythraea produces the commercially important macrolide
antibiotic, erythromycin A, in addition to an unidentified red pigment from which the organism obtained it name (Labeda, 1987). Hsieh and Kolattukudy (1994)
investigated the effect of malonyl-CoA decarboxylase disruption on erythromycin production. They observed that the mutant strain lost the capacity to produce erythromycin and a red pigment, while the S. erythraea CA 340 parent strain produced erythromycin and ‘became red as the culture grew dense’. They also reported that upon addition of exogenous propionate to the growth medium, the mutant was able to re-establish erythromycin production but was not able to restore red pigment production. Clark et. a l (1995) examined the effect of controlled oxygen limitation on secondary metabolite formation and observed that S. erythraea wt NRRL 2338 produced an ‘unidentified soluble red pigment’ under oxygen-limited conditions.
Recently, the genes involved in the production of a red pigment produced by S. erythraea wt RV have been identified and characterized (Cortes et a l, 2002). Their work also shows that a variety of pigments derived from tetrahydroxynaphthalene are produced. Consistent with these results, Udea et. a l (1995) identified a water-soluble red-pigment produced by S. griseus and determined that the 2 genes rppA and rppB were necessary for pigment production (Ueda et a l, 1995). Funa et. a l (1999) identified the red-pigment product of rppA produced by S. griseus as 1,3, 6, 8- tetrahydroxynaphthalene (THN). They proposed that a chalcone synthase-like enzyme incorporates five malonyl-CoA units into a pentaketide before cyclizing to THN, which can be oxidized to form flaviolin. Flavolin can form various pigmented compounds which was evident when rppA was subcloned into E. colt a mixture of
Ph .D . Thesis M e ta b o lic E n g in eerin g o f S. tryihrzQ i.
pigmented compounds was produced (Funa et a l, 1999). Even though we have not yet structurally characterized the pigments discussed in this report, it is possible that they are similar or identical to the compounds reported above derived from THN. However, very recently in a poster presentation, Weber and colleagues (2002) found that one of the red pigments produced by S. erythraea was pyomelanin, a shunt product o f tyrosine metabolism. This result presents a new possible metabolic route to red pigment biosynthesis to compare to the THN hypothesis.